dumplingai-transcripts-media

当用户想要通过DumplingAI获取转录、字幕、语音文本提取或媒体到文本的工作流程时。此外，如果用户提到YouTube转录、TikTok转录、转录这个视频、提取字幕、从音频中提取文本、从URL获取转录，或者希望在根据视频内容进行写作或总结之前检查其内容，也使用此功能。每当任务是首先将视频或音频转换为文本时，请使用此功能。对于将YouTube转录转换为完成的文章，请参阅youtube-to-blog-post。

DumplingAI Transcripts And Media

Allowed Commands

dumplingai catalog search <prompt>
dumplingai catalog details <type> <id>
dumplingai run <type> <id> --input '<json>'

Workflow

Convert the request into a short keyword query such as youtube transcript, tiktok transcript, or extract video.
Run dumplingai catalog search "<job>".
Inspect the best candidate with dumplingai catalog details <type> <id>.
Execute with the media URL or supported input payload.
Save long transcript payloads to .dumplingai/ and inspect only the relevant fields.
If the user wants a blog or newsletter draft from the transcript, hand off to youtube-to-blog-post.

Search Phrases

youtube transcript
tiktok transcript
extract video
transcription
audio transcript

Output Strategy

dumplingai catalog search "youtube transcript" > .dumplingai/catalog-search.json
dumplingai catalog details capability get_youtube_transcript > .dumplingai/youtube-transcript.json
dumplingai run capability get_youtube_transcript --input '{"videoUrl":"https://youtube.com/watch?v=ID"}' > .dumplingai/transcript.json
head -80 .dumplingai/transcript.json
rg '"transcript"|"text"|"segments"|"captions"' .dumplingai/transcript.json